Ladies and gentlemen, fans of generated images, I would appreciate your feedback, as well as uploaded images created using this model...
This checkpoint was created for generating film photographs. This model contains trained LORA, color photographs of past years for maximum realism.
Recommended settings for generation:
Ver.2.
photo RAW,Incredibly {prompt}, exceptional shallow depth-of-field capabilities, atmospheric blur,vivid colors,, high quality textures of materials, volumetric textures, coating textures, metal textures, dusty atmospheric haze
Ver.1.
Fine art photography, (amateur photo), (analog photo:1.2,film grain),({prompt}),vivid colours,realistic,photographic, (natural colors, correct white balance, color correction, dehaze,clarity), background rich in details, conceptual image,masterpiece, award winning photography
Steps: 28, Sampler: DPM++ 2M SDE Karras, CFG scale: 6, ADetailer, Hires prompt: Hires upscale: 1.35, Hires steps: 14, Hires upscaler: 4x_NMKD-Siax_200k
Description
FAQ
Comments (4)
I wanted to like this model, but what can I do with this:
Error: Could not load the stable-diffusion model! Reason: 'Attention' object has no attribute 'to_to_k'?
Same problem. Altough it started to appear after I have updated SDNEXT. Before using newer version there was no errors.
It's a pretty good model, I liked it. Great at artistic styles and quite flexible. But in some difficult prompts, image starts breaking and artifacting.
Impressive checkpoint. Beautiful images, and colors are very vibrant. Thank you!